The Role of Extracorporeal Membrane Oxygenation Simulation Training at Extracorporeal Life Support Organization Centers in the United States.

INTRODUCTION: Extracorporeal membrane oxygenation (ECMO) requires a multidisciplinary healthcare team. The Extracorporeal Life Support Organization publishes training guidelines but leaves specific requirements up to each institution. Simulation training has shown promise, but it is unclear how many institutions have incorporated simulation techniques into ECMO training to date. METHODS: We sent an electronic survey to ECMO coordinators at Extracorporeal Life Support Organization sites in the United States. Participants were asked about training practices and the use of simulation for ECMO training. Descriptive results were reported as the percentage of total responses for each question. Logistic regression was used to identify characteristics associated with simulation use. RESULTS: Of 94 responses (62% response rate), 46% had an ECMO simulation program, whereas 26% report a program is in development. Most (61%) have been in operation for 2 to 5 years. Sixty-three percent use simulation for summative assessment, and 76% have multidisciplinary training. Access to a simulation center [odds ratio (OR) = 4.7, 95% confidence interval (CI) = 1.7-12.5], annual ECMO caseload of greater than 20 (OR = 2.5, 95% CI = 1.5-5.8), and having a pediatric cardiothoracic intensive care unit (OR = 2.8, 95% CI = 1.2-6.7) are each associated with increased likelihood of mannequin-based ECMO simulation. Common scenarios include pump failure (93%), oxygenator failure (90%), and circuit rupture (76%). DISCUSSION: Extracorporeal membrane oxygenation simulation is growing but remains in its infancy. Centers with access to a simulation center, higher caseloads, and pediatric cardiothoracic intensive care units are more likely to have ECMO simulation programs. Extracorporeal membrane oxygenation simulation is felt to be beneficial, and further work is needed to delineate best training practices for ECMO providers.

The effect of multimedia replacing text in resident clinical decision-making assessment.

Multimedia in assessing clinical decision-making skills (CDMS) has been poorly studied, particularly in comparison to traditional text-based assessments. The literature suggests multimedia is more difficult for trainees. We hypothesize that pediatric residents score lower in diagnostic skill when clinical vignettes use multimedia rather than text for patient findings. A standardized method was developed to write text-based questions from 60 high-resolution, quality multimedia; a series of expert panels selected 40 questions with both a multimedia and text-based counterpart, and two online tests were developed. Each test featured 40 identical questions with reciprocal and alternating modality (multimedia vs. text). Pediatric residents and rising 4th year medical students (MS-IV) at a single residency were randomized to complete either test stratified by postgraduate training year (PGY). A mixed between-within subjects ANOVA analyzed differences in score due to modality and PGY. Secondary analyses ascertained modality effect in dermatology and respiratory questions using Mann-Whitney U tests, and correlations on test performance to In-service Training Exam (ITE) scores using Spearman rank. Eighty-eight residents and rising interns completed the study. Overall multimedia scores were lower than text-based scores (p = 0.047, η p2  = 0.04), with highest disparity in rising interns (MS-IV); however, PGY had a greater effect on scores (p = 0.001, η p2  = 0.16). Respiratory questions were not significantly lower with multimedia (n = 9, median 0.71 vs. 0.86, p = 0.09) nor dermatology questions (n = 13, p = 0.41). ITEs correlated significantly with text-based scores (ρ = 0.23-0.25, p = 0.04-0.06) but not with multimedia scores. In physician trainees with less clinical experience, multimedia-based case vignettes are associated with significantly lower scores. These results help shed light on the role of multimedia versus text-based information in CDMS, particularly in less experienced clinicians.

Glucose variability and survival in critically ill children: allostasis or harm?

OBJECTIVES: To assess whether individual blood glucose variability in critically ill children is associated with increased mortality and to define the temporal patterns of blood glucose variability during critical illness in children. DESIGN: Retrospective cohort study. SETTING: A 20-bed pediatric intensive care unit in a children's hospital. PATIENTS: Patients aged 0-20 yrs and with at least 12 blood glucose measurements taken within the first 72 hrs of pediatric intensive care unit admission. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: A total of 101 eligible patients had 3,144 measured blood glucose concentrations with 16% mortality. Nonsurvivors had higher median blood glucose concentrations (129 mg/dL vs. 118 mg/dL, p < .01), more hyperglycemia (blood glucose >200 mg/dL) (88% vs. 59%, p < .05), and more hypoglycemia (blood glucose <60 mg/dL) (56% vs. 15%, p < .01) than survivors. The mean blood glucose range (257 mg/dL vs. 185 mg/dL, p < .01) and the blood glucose variability (63 mg/dL vs. 45 mg/dL, p = .02) were greater in nonsurvivors compared with survivors. Blood glucose variability tertiles were proportionately associated with increasing mortality: 6% vs. 15% vs. 27% (p = .07). Compared with survivors, daily blood glucose variability was significantly higher in nonsurvivors during the first 48 hrs of admission and after 1 wk of admission. After controlling for confounders, individual blood glucose variability was associated with higher pediatric intensive care unit mortality for each mg/dL of blood glucose concentration (adjusted odds ratio, 1.03; 95% confidence interval, 1.01-1.05). CONCLUSIONS: Glucose variability is common in critically ill children and is associated with increased mortality. Whereas early alterations in blood glucose may represent allostasis, later fluctuations in blood glucose may represent an alteration of autoregulation with resulting higher mortality. Control of variability may need to be incorporated into glycemic control regimens.